Welcome to Computational Musicology 2025!
Welcome to Computational Musicology 2025! This is going to be the intro page for the final portfolio
Visualising the AI Song Contest
Exploring Tempo Distributions This density plot visualizes the distribution of tempos in the AI Song Contest dataset. The x-axis represents tempo (beats per minute, BPM), while the y-axis shows density, indicating how frequently different tempo ranges appear in the dataset.
The fill color represents danceability, meaning that more danceable songs are highlighted in different shades. This allows us to explore whether certain tempos are more common in highly danceable tracks.
Key Insights: The dataset shows clear peaks, suggesting that AI-generated songs tend to favor specific tempo ranges. The smooth distribution helps identify whether AI compositions cluster around conventional musical tempo patterns or diverge from human-created norms. By incorporating danceability, we gain additional insight into the relationship between rhythmic elements and listener engagement. This visualization helps us understand how AI-generated music organizes rhythmic elements, potentially revealing biases or preferences in algorithmic composition.
Discussion: Insights & Next Steps
- Patterns observed: The dataset shows clear peaks, suggesting that AI-generated songs tend to favor specific tempo ranges. The smooth distribution helps identify whether AI compositions cluster around conventional musical tempo patterns or diverge from human-created norms.
- Limitations: The dataset lacks information on lyrical content and emotional perception.
My Tracks & Creation Process
This repository also includes original compositions. These tracks were created using a combination of digital audio workstations (DAWs), MIDI synthesis, and AI-assisted composition tools. The goal was to experiment with algorithmic composition techniques while maintaining human musical intuition.